The Cliff at the Border1

Lant Pritchett

I begin with Keynes famous description of the world (well, London, well, upper
middle class London) at the apex of the first globalization just before that worlds
tragic end in the carnage of World War I, dividing his text into four sections:
What an extraordinary episode in the economic progress of man that age
was which came to an end in August 1914! in[to] the middle and upper
classes, for whom life offered, at a low cost and with the least trouble,
conveniences, comforts, and amenities beyond the compass of the richest
and most powerful monarchs of other ages. The inhabitant of London
could order by telephone, sipping his morning tea in bed, the various
products of the whole earth, in such quantity as he might see fit, and
reasonably expect their early delivery upon his doorstep;
What better description could there be of the benefits to the sovereign consumer of the
liberalization of goods (note already his awareness of the role of income inequality in
this, these are available to the middle and upper classes). This description is even truer
today, as technical advance has put even more conveniences, comforts and amenities
like air travel, cell phones, medical care, air conditioningon offer and globalization of
the trade in goods has augmented thisleading to goods of amazing variety (e.g. foods
from every corner of the earth), availability (e.g. fruits year round), and low cost. The
material lifestyle of the middle class of rich countries today far exceeds that of the
nobility of centuries ago.
he could at the same moment and by the same means adventure his
wealth in the natural resources and new enterprises of any quarter of the
1

This is a contribution to the volume, Equity and Growth in a Globalizing World,

edited by Ravi Kanbur and Michael Spence and is based, very loosely, on comments at a
session sponsored by the Growth Commission on globalization. I would like to thank
Michael Clemens for comments on a preliminary version.

world, and share, without exertion or even trouble, in their prospective

fruits and advantages; or he could decide to couple the security of his
fortunes with the good faith of the townspeople of any substantial
municipality in any continent that fancy or information might recommend.
Again, a wonderful description of the glories of the liberalization of capital
interestingly both equity and debt, and with more extensive bond markets that even exist
today (the townspeople of relatively few substantial municipalities are able to issue
bonds internationally, or so much more prosaically, invite people to couple their
fortunes with their good faith). But again today, one can, with the click of a button,
adventure [your] wealth into index funds of Indian stocks or Brazilian bonds.
He could secure forthwith, if he wished it, cheap and comfortable means
of transit to any country or climate without passport or other formality,
could dispatch his servant to the neighboring office of a bank for such
supply of the precious metals as might seem convenient, and could then
proceed abroad to foreign quarters, without knowledge of their religion,
language, or customs, bearing coined wealth upon his person, and would
consider himself greatly aggrieved and much surprised at the least
interference.

Often overlooked in favor of the much more widely cited passages about Londoners
enjoying the products of the earth, this is an excellent description of the mobility of
people. Notice that the travel is without passport or other formality. Notice also the
sensitivity (or lack thereof?) to the inter-personal distribution of income and to who these
benefits of globalization are available tucked into just how one gets foreign exchange for
travel: one dispatches ones servant of course.
This liberality has not been recreated, in two senses. Even for the elite of the
world there is not longer travel without formalityeven as a traveler from the most
powerful nation on earth and with sufficient funds, I can attest to the need for constant
formalities to travel. But more importantly, at least within certain areas there has also

been free mobility for all peopleup until roughly 1914 there were open borders for the
movement of labor from Europe to not just the Americas, Australia, and New Zealand but
also to Latin America, and for the citizens of Great Britain to other British Colonies, and
more complex flows (e.g. some voluntary, some restricted, some forced) elsewhere
within the Empire as well (e.g. the movement of Indians to Africa and the Caribbean)
These three passages illustrate the first point I wish to make. The world of the
first great globalization came to an end, or at least the beginning of its end, in August
1914. As we all know, the end of this first globalization was followed by some extremely
nasty bits of history, with two extremely bloody world wars, the rise of
Leninism/Stalinism in Russiawith brutality and famines and a staggering loss of life
orchestrated by the state--the rise of Fascism in Europewith the attempted genocide of
Jews, a staggering loss of life orchestrated by the state.
Keynes himself, conscious of the world having lost the first peace, was
instrumental in attempting to win the second peace by establishing institutions following
World War II to recreate the globalization that has created the previous extraordinary
episode in the economic progress of man and avoid calamities. In this, the world has
been fantastically successful in recreating two of the three liberalities. But in the 30 years
between 1914 and 1944 (the Bretton Woods conference) apparently all appetite had
disappeared for the third element of the first great globalization. There was no attempt to
recreate the globalization of labor markets, no creation of institutions of encourage and
manage that process, no equivalent of the WTO or IMF for the movement of people2.

There was in fact an International Organization for Migration set up (which still exists) but the objective
of that group was to facilitate the return of refugees, rather than having the broader policy agendas of the
institutions intended at Bretton Woods (the WTO taking more than 50 years to acquire independent
organizational status).

This, combined with the movement for de-colonization following World War II has led
us all into the grand experiment I call the world of the POSEBLL, a Proliferation of
Sovereigns combined with Everything But Labor Liberalization. This is a sufficiently
ugly acronym I alert the reader why I use it: to pose a question with a terrible pun at the
end: Is more than the POSEBLL possible?
The first point I wish to make is that the world of the POSEBLL has led, as
expected, to the equalization of prices of goods, equalization of the prices of capital. But,
perhaps unexpectedly, it has also led to very uneven progress in the newly proliferated
sovereigns and this, combined with binding quantitative restrictions on the movement of
labor, has also led to massive gaps in the wages of equivalent labor around the world and
sustained divergence in the per capita incomes across nation-states.
Keynes goes on to make a much deeper point about attitudes, which leads to the
second point I wish to make.
But, most important of all, he regarded this state of affairs as normal,
certain, and permanent, except in the direction of further improvement,
and any deviation from it as aberrant, scandalous, and avoidable. The
projects and politics of militarism and imperialism, of racial and cultural
rivalries, of monopolies, restrictions, and exclusion, which were to play
the serpent to this paradise, were little more than the amusements of his
daily newspaper, and appeared to exercise almost no influence at all on
the ordinary course of social and economic life, the internationalization of
which was nearly complete in practice.
What is interesting about this passage (and the end of the one above) is that the
internationalization was regarded as a perfectly normal course of events and one would
have been aggrieved and surprised at any attempts to deviate from this obvious and
natural pattern of free movement (at least of Londoners with servants). Moreover, he
regarded this situation as certain and permanent the logic of internationalization was

such obvious common sense it was impossible to conceive a move to a fundamentally

different arrangement. But this world did end, dramatically, and for a very long time, as
what was so certain about the world became first contested and then obviously false.
Since I have begun with quotes from Keynes, let me turn to his great predecessor,
Karl Marx:
Hegel remarks somewhere[*] that all great world-historic facts and
personages appear, so to speak, twice. He forgot to add: the first time as
tragedy, the second time as farce. Men make their own history, but they
do not make it as they please; they do not make it under self-selected
circumstances, but under circumstances existing already, given and
transmitted from the past. The tradition of all dead generations weighs
like a nightmare on the brains of the living.

The second globalization, the world of the POSEBLL, is repeating the great worldhistoric fact of the first great globalization, but this time, following the tragedy of its
demise, as farce. The real puzzle is why people continue to assert that we live in an age
of globalization when we so obviously do not live in a world of globalization. We live in
an age of nationalization, a nationalization which is a deep, radical, and unprecedented in
the long history of mankind3. After all, your first encounter in every country you visit is
with the people who enforce the regulations about the movement of people intended to
keep the world from being flatand it is obviously successful (which I will document at
length with new empirical results about wage gaps).
The second section of this paper addresses the conundrum that in our supposedly
globalized world we think about equity in completely nationalized ways. This
3

I return to this point, especially in the second section. What is unique is not that the world is divided into
a large number of sovereign states, this process of globalizing empires followed by fracturing into
independent states has been repeated a number of times. What is new this time around is the association of
the relatively new concept of the nation with the age-old concept of the state, which is what I refer to as
nationalizationwhich is the combination of statehood and an ideology of the nation into nationstates.

nationalization of our lived reality is so deep and so complete that lists of issues of
equity and justice in a globalized world will include: (1) the differentials between
Guatemalan men and Guatemalan women are raised as an example of gender inequity,
(2) the treatment of indigenous Guatemalans versus other Guatemalans is raised as an
example of ethnic inequity, (3) the issue of the gap between the landed and landless is
raises as an example of persistent economic inequity, (4) the reduction in trade barriers
causing an increase inequality across people of different skills is raised as an example of
the inequities induced by globalization, (5) the issue of education gaps between rich
and poor within Guatemala are presented as examples of perpetuation of
poverty/inequality across generations, (6) the issue of value chains in coffee sold by
Guatemalan farmers are adduced as inequities in globalization, etc. One can get very far
into lists of equity problems around which people are willing to mobilize before the
gap in wages induced by US restrictions on the movement of Guatemalans ever comes
up. Thomas Friedman can write that the world is flat in the same way, and for the same
reasons, that Thomas Jefferson could write that all men are created equal while owning
slavesthe nationalizing tradition of dead generations weighing on our (collective)
brains.
In discussions of globalization and equity I am a triply impolite guest. First, I
dispute the premise that, from the point of view of the developing countries,
globalization is a primary phenomenon of our time. Second, I argue that all of the
issues discussed under the context of globalization and equity around everything but
labor liberalization are trivial compared to the one that is notcross-border flows of
labor. Third, I argue that we lack a coherent way of talking about equity that does not

depend on an arbitrary advancement of nationality to a first rank justification of

acceptable differences in well-being, which is an incoherent way of talking about
globalization.
I) The Proliferation of Sovereigns
I was born in Idaho, a state in the upper Northwest primarily famous for potatoes
and, at least for a while, for crazy white supremacists and survivalists. Suppose that in
2009 Idaho withdrew from the (formerly) United states and became a sovereign
country, with its own flag, military, money, laws, courts, passports, etc.all of the
trappings of a sovereign state. At independence suppose Idaho also simultaneously
announced a 25 percent tariff on all goods entering Idaho from any foreign country,
including the remaining other 49 states. Now suppose times goes on and ten years later
in 2019 Idaho liberalizes its trade by reducing their tariff to 10 percent. One could
easily find the impact of this policy shift of some interest. But any academic who
proposed that integration of Idaho into the US economy was the primary question of
interest due to this modest liberalization of cross-border flows would be laughed out of
the room. Obviously the key experience of interest would be the disintegration of Idaho
from the rest of the United States from the acquisition of sovereignty, not the subsequent
liberalization.
A key feature of the post-WWII period is the incredible proliferation of sovereign
states so that the number of nation-states has risen from around 50 to around 200. As
illustrated in Figure 1, this happened in roughly three waves, a group immediately
following WWII (including importantly Indonesia and India and Pakistan), a group of
primarily African countries gaining independence around the early 1960s (with a another

group of the formerly Portuguese colonies gaining independence in 1974/5), and finally
the proliferation from the disintegration of the Soviet Union and some of its satellites
(e.g. Czechoslovakia, Yugoslavia). So, while all of the major industrial countries have
been sovereign states for 100 years or more, most poorer countries (the obvious regional
exception being Latin America) were in some kind of colonial or quasi-colonial
relationship that limited sovereignty over policies until (relatively) recently.
Figure 1: Number of new states added per year, 1943-1994 by region
20

18

16

ssa

mena

eca

lac

eap

other

14

12

10

1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994

Many of these countries, on acquiring sovereignty over economic policy did

pursue a rather aggressive form of the use of trade barriers for many reasons (including
revenue needs from tariff collections and export taxes in the absence of other instruments,
reaction to the forced liberality under colonialism and the wish to assert autonomy, as
well as a general ideology of state led industrialization). Since the emergence of the debt
crisis (signaled by Mexicos August 1982 announcement it would not be able to service

its debt) there has been a cumulative, gradual, but by now nearly complete shift towards a
dismantling of the more egregious barriers and a general liberalization of trade.
However, very few countries have gone beyond the liberalization of the crossborder trade in goods (including regional customs unions and free trade areas) and the
relaxation of control on some types of capital flows, to engage in any really significant
deep integration (e.g. common currencies, free movement of labor, harmonization of
regulation). In fact, about the only significant experience with deep integration has
been that of the European Union and, to a much lesser extent, NAFTA.
Which is more important in the experience of say, Kenya or Jamaica or Indonesia,
the nationalization implicit in sovereignty or the globalization implicit in cross-border
liberalization of the flows of goods (and perhaps capital)?
A set of recent studies examined the extent to which trade between Canadian
provinces exceeded that of trade between a Canadian province and a US state. The
pioneering paper by McCallum (1995) showed that trade between British Columbia and
Ontario was US$1.4 billion while trade with Texas was only 155 million, whereas if one
predicted the trade with Texas if it were a province of Canada and hence there was a zero
border effect would be US$2.1 billion. It is striking that one of the most liberalized
borders in the world still appears to be a huge deterrent to trade. This suggests that the
mere fact of the borders createdalong with the creation of different currencies,
difference courts, etc.by the proliferation of sovereigns may have been only mildly
mitigated by the subsequent liberalizations to date.

Figure 2: Differences in trade flows between British Columbia comparing

Ontario and Texas demonstrates the importance of the US-Canada border
2.5
2.1

Billions

2
1.5

Ontario, Canada
1.4
Texas

1
0.5
0.155

Predicted BC exports
to Texas without a
border

0
British Columbia's exports to:

Source: McCallum, 1995

While there are unquestionably increased cross-border flows of goods in most
countries and in capital in some, the question is whether this implies the world is, in any
relevant sense, globalized or that globalization is a useful lens to examine recent
times. I argue that at the very least this is not obvious, especially for the poorer countries
which acquired sovereignty and hence acquired a set of institutions (e.g. distinct
currencies, domestic regulations, independent courts) that, no matter how liberal
explicit trade policy is, create frictions to trade.
II) From the top of the cliff for labor you cannot see other globalization issues
That there are gaps in real wages, adjusted for purchasing power, across countries
is obvious. These need not be due to restrictions on the movement of labor or lead to
pressures for labor movement if the differences are due to the intrinsic productivity of the
worker. After all some basketball players make millions of dollars, much more than I do,

yet this, in and of itself, it no evidence that there are barriers to my employment as a
professional basketball player other than that fact that, being both short and slow, I have
low productivity in that occupation. But there clearly are restrictions that are intended to
restrict the mobility of persons across nation-state bordersevery airplane arriving in the
US is met by (armed) officials whose job it is to prevent the entry of unauthorized
persons. The question is, how much more would people from other countries make if
they did not face these restrictions and people in the USA could pay them a wage that
reflected their productivity in the USA? Or, more prosaically, how high is the cliff that
blocks labor at our borders?
A recently completed study of mine, with two co-authors, Michael Clemens and
Claudio Montenegro, answers precisely this question. We take advantage of two sources
of data. First, we used a collection of data sets from around the world that recorded
wages of individuals and some relevant characteristics (e.g. their years of schooling, age,
sex). Using this data we could adjust the wages that individuals make in their home
country for observable characteristics related to their productivity. The second source of
data was the US Census. The US Census also collects information on wages and the
characteristics of individuals such as years of schooling, age, etc. Most importantly, it
also contains information on a persons country of birth and when they arrived in the
USA.
Let me use Peru as an example to illustrate what these two sources of data allow
us to do. We can compare the predicted real consumption (PPP adjusted) wages of a
Peruvian born, Peruvian educated with 9 years of schooling, male, 35 years old, urban
resident working in Peru to the observable equivalent personPeruvian born, Peruvian

educated with 9 years of schooling, make, 35 years old, urban resident. Figure 4 just
illustrates that this involves estimating a wage profile, the relationship between wages
and characteristics (illustrated for just one characteristic, X, but in reality this is a multidimensional surface), in Peru and a wage profile for late-arriving Peruvians in the USA.
Then one can drill down through those wage surfaces at any given point on the wage
profile to estimate the wage gap between observably identical individuals in the USA and
in Peru. The figure is drawn with squiggly lines to emphasize that the empirical
procedure imposes almost no assumptions on the shapes and form of the two profiles
(e.g. we do not impose the wage returns to schooling are the same in the two countries).
Figure 4: Wage profiles to estimate the wage gap

(Hypothetical) estimated wage

profile of Peruvian born late
arrivers working in the USA

Real
Wages
A

Wage ratio=A/B,
Wage Gap=A-B
B

(Hypothetical) estimated wage

profile of Peruvian born working
in Peru

Characteristic X (e.g.
Schooling, Age)

Given our data, we were able to estimate the wage ratios of observably equivalent
workers in the US and 42 different developing countries around the world. The results

for our particular comparison are in Table 1. The apparently same worker from these
countries makes five times as much in the USA as in their home country. That is, on
average an annual wage income P$15,000 higher.

Table 1: Wages of observably equivalent workers across the US border for 42 countries
Comparing wages with college
degree in home to workers with
primary schooling working in
the USA

Comparing low skill (9 years of schooling), male, 35 year old

Country

Annualized wage difference

Ratio

Yemen

$21,772

15.45

11.43

Nigeria

$17,155

14.85

7.79

Egypt

$18,660

11.92

11.93

Haiti

$15,738

10.31

4.19

Cambodia

$20,737

7.45

6.4

Sierra Leone

$15,977

7.43

3.7

Ghana

$17,164

7.12

4.22

Indonesia

$17,478

6.72

3.17

Pakistan

$18,019

6.57

2.95

Venezuela

$17,471

6.57

3.69

Cameroon

$17,807

6.53

7.38

Vietnam

$16,753

6.49

3.92

India

$19,340

6.25

2.96

Jordan

$16,439

5.65

3.98

Ecuador

$14,300

5.16

3.26

Bolivia

$15,455

5.03

3.34

Sri Lanka

$14,666

4.95

1.26

Nepal

$11,524

4.85

4.37

Bangladesh

$14,891

4.60

2.19

Uganda

$15,318

4.38

2.3

Ethiopia

$14,772

4.35

2.4

Guyana

$16,888

3.87

1.39

Philippines

$13,615

3.82

1.42

Peru

$15,149

3.79

1.6

Brazil

$17,423

3.76

1.66

Jamaica

$15,421

3.63

1.55

Chile

$16,057

3.53

1.6

Nicaragua

$13,412

3.52

1.42

Panama

$14,368

3.36

1.54

Uruguay

$20,962

3.10

1.9

Guatemala

$12,295

2.94

1.73

Colombia

$12,330

2.88

1.65

Paraguay

$17,674

2.78

1.1

South Africa

$20,311

2.75

0.65

Turkey

$12,877

2.68

1.46

Argentina

$13,700

2.54

1.37

Mexico

$10,679

2.53

1.31

Belize

$14,959

2.43

1.16

Thailand

$9,859

2.17

1.04

Costa Rica

$9,982

2.07

1.24

Morocco

$8,970

2.00

0.62

Dominican Rep.

$8,912

1.99

1.3

Mean

$15,411

5.11

2.99

Median

$15,438

4.11

1.82

Of course, the correcting the wages for a few simple observable characteristics
might not fully adjust for equal productivity such that we know that the gains in the table
above would be the gains to a worker moving across the border. Perhaps comparing the
wages of Peruvians here to Peruvians there, even correcting for their education,
overstates the wages gains from migration because those that moved would have been
more productive had they remained at home than those that did remain at home. That is,
perhaps people who move have more pluck or drive or ambition or some other personal
characteristic that makes them more productive in either place. In this case positive
selection of migrants would lead to an overstatement of the wage gain from just
comparing observable equivalent individuals. In our paper, CMP (2008) we devote
considerable attention to this issue and deploy several new sources of evidence about the
range of the magnitude of the adjustment for positive selection we should typically
expect (while anticipating it would vary across countries). Overall we find that there is
some evidence of positive selection, such that, if one wanted to be conservative about the

adjustment you could divide the ratios of the wages observably identical workers by a
factor of around 1.2 (the range is from 1.0 to 1.4) to get to equal productivity workers.
Even after adjusting for the potential positive selection of migrants, this still
suggests that the gain to a low-skill worker from these 42 countries of moving across the
US border is to increase wages by a factor of 4.26, for a gain of around P$13,000 per
year.
Now, it is possible that even this gain is because people are sufficiently wedded to
their own place, language, culture, family and social ties that these wage gains are not
worth the other real and psychic costs of moving. Fortunately, to address this question
we have a near natural experiment in that Puerto Rico has remained a territory of the
United States and as such, its citizens have the right to live and work in the USA. So we
can do exactly the same procedure for Puerto Rico and ask what is the wage ratio for
observably equivalent workers with spatially disintegrated, linguistically and culturally
distinct places when there are no obstacles to labor movement? Puerto Ricos wage
ration is 1.5, which given the very long period for which labor movement has been free
might be near a sustained equilibriumthat wages would have to be 50 percent higher to
induce workers to move to a foreign country.
Of course, this is enormously higher than the observed ratios for other Central
American and Caribbean countries. Dominican Republic has the lowest observed wage
ratio, only 2, but most other Central American countries have ratios almost twice as
highJamaica is 3.6, Guyana is 3.8the median for Central American and Caribbean
countries is 2.94, almost exactly twice the ratio for Puerto Rico. This suggests that the

observed wage differentials are the result of the border controls, not because of any lack
of interesting in taking advantage of the wage gains.
Not surprisingly, this evidence suggests that for labor movement the border
matters, a lot. Even if we discount the estimates of wage differences of observably
equivalent workers by a factor of 1.5 to adjust for selection and for costs of moving, the
gains in wages to a low skilled worker are P$10,000.
Not only is the world not flat, it is not a curb, not a barrier, but a massive cliff at
the US border (and, one suspects, similarly sized cliffs for most other rich countries).
There are lots of other issues that are discussed in the context of globalization and
equityincluding the movement of capital, the effects of the liberalization of trade,
creating anti-poverty programs in poor countries, the working conditions of sweatshop
workers, etc. However from the top of the cliff of labor restrictions all of these are barely
visible.
One way of making these comparisons is ask: how long we would have to let a
worker work at their market wage in the USA (just meaning there is a voluntary
employer willing to pay them based on their productivity) in order to be equal to the
benefits from a lifetime of benefiting from other programs or interventions.
For instance, the provision of micro-credit had garnered enormous amounts of
attention, with on the of pioneers in revivifying micro-credit, Muhmmad Yunus, winning
a Nobel Peace Prize for his efforts. So, lets ask, how many weeks would a Bangaldeshi
man have to work in the USA in order to produce a gain equal to a generous estimate of
the gain in net present value of a lifetime of access to micro-credit. A pioneering (if
controversial because it is higher than others) estimate by Pitt and Khandker (1998) of the

net return on micro-loans to Bangladeshi women is 18%. Taking this substantial return at
face value, this translates into an increase in annual household income of US$65 at
purchasing power parity, so that a lifetime of continuous access to lending with these
returns would return US$683 in net present value.4
From Table 1 above an observational equivalent low-skill Bangladeshi man
makes P$14,891 more a year in the USA. To be conservative, scale this figure back by
1.5 to account for positive selection and psychic costs of moving. At that level he would
have to work four weeks in the United States to have a gain in income equal to a lifetime
of micro-credit5.

Obviously one would have to add a few weeks to pay transport costs

and some for expenditures while in the USA, but a single seasonal access of three months
could provide savings more than equal to the total gain from a lifetime of micro-credit.
Table 2: A comparison of annual wage gains from international movement of the
marginal worker, to present-value lifetime wage gains to the marginal worker from
different in situ antipoverty interventions

Intervention

Microcredit
Anti-sweatshop
activism
Additional year of
schooling (at
zero cost)
Deworming

Country

Bangladesh

Present-value
lifetime wage
increment due to
intervention (US$
at PPP)

Annual wage
increment due to
working in US
(US$ at PPP)

Weeks of US work
equivalent to
lifetime NPV of
intervention

700

~10,000

Indonesia

2,700

~12,000

30

Bolivia

2,250

~11,000

11

Kenya

71

~11,500

0.3

Source: CMP (2008a).

4

Their estimate of the return to males is 11%, but we use the higher figure for females to be conservative. Average
annual female borrowing is Tk3415, or US$361 at PPP using the average PPP conversation factor from World Bank
(2007) over the relevant period (1986-1992) of 9.47. The resulting increase in household income is thus Tk615 or
US$65 at PPP. Average life expectancy in Bangladesh during 1986-1992 was 55 years and average borrower age in the
sample is 23, and a 33-year stream of US$65 payments (including one at time 0) discounted at 10% has a net present
value of US$683. At 5% the value is US$1091 and at 15% it is US$493.
5

I am aware of all the problematic aspects of gender implied by income accruing to a man versus a woman,
but the regressions were used for men since labor force participation is so much more complex an
econometric issue for women.

We have done similar calculations for other (anti)-globalization or anti-poverty

initiatives intended to address equity. For instance, anti-sweatshop activism does
appear to have led to wage gains for Indonesian workers, such that a low skilled
Indonesia worker would have to work half a year in the USA to be equal to a lifetime of
the gains from anti-sweatshop movement. Expanding schooling, a policy that spawns
social movements, international resolutions, MDG commitments, and the like, does
produce wage gains. To produce the equivalent of the gross lifetime gains from an
additional year of school (not even taking out the opportunity cost) a Bolivian worker
would have to work in the USA for 11 weeks.
One could multiply these examples, which all hinge on the same basic simple, but
inexorable arithmetic: most gains from in-situ anti-poverty interventions are measured in
percents of local income while the gains from labor mobility to the USA are measured in
factor multiples so the annual gains from labor mobility are typically two orders of
magnitude larger than even the most optimistic estimates of development actions.
This is not of course to pose these are alternatives, one could easily both expand
micro-credit and reduce the barriers to labor mobility. The point is that, at the margin,
the gains to poor people from relaxing existing barriers to labor mobility are enormous
relative to everything else on the development table so doing one out of concern for
equity of globalization and not advocate the other makes almost no sense6.
This same logic also applies to the potential gains from further liberalization of
the already quite liberalized markets for goods or capital.

These calculations are even more dramatic if one factors in costs, as micro-credit or a year of schooling
for instance cost real resources while expanding migration, by most estimates, has almost zero welfare cost
(and in many instances, substantially positive gains) for the receiving country so it is at least potentially
win-win.

Caselli and Feyrer (2006) calculate the marginal product of capital (MPK) across
countries. They find that marginal products of capital across countries are essentially
equalized. In fact, by their estimates, which correct the nave estimates of the
marginal product of capital for differences in natural capital and in the price of output,
the return to capital is lower in poor countries than rich countries (8.4% vs. 6.9%, from
Table 2). If this research is to be believed the gains from facilitating capital flows to poor
countries are modest. Since the marginal product of capital is so nearly equalized, their
estimate of the welfare gains from complete equalization of the MPK across countries is
only one tenth of one percent of world GDProughly $65 billion. Even if one went
beyond liberalization and subsidized the flows of capital, the net gain, the difference
between the financing cost and MPK limited by the low MPK.
While borders may create substantial barriers to trade, the gains from
liberalization, the reduction in trade barriers are, by now, quite modest. An estimate by
the World Bank 2005, p. 128) is that an elimination of all remaining policy barriers to
trade worldwide, would produce welfare gains to the developing countries of roughly
$109 billion in annual income by 2015.
In contrast to these modest gains from further liberalizations of goods or capital
markets the estimates of the gains from the fanciful counter-factual of a complete
liberalization of labor mobility are that world GDP would roughly double7 At current
levels of GDP this implies gains of 65 trillion dollars, roughly three orders of magnitude
larger than the world gains from MPK equalization or than the developing country gains
from all remaining trade liberalization. Or, rather than even entertaining the borderline
7

Hamilton and Whalley (1984) estimate was a rough doubling of world output per person. Klein and
Ventura (2004) use a calibrated general equilibrium model with capital mobility and estimate gains
between 94% and 172%.

facetious estimate of open borders one can calculate the gains from a modest relaxation
of the constraints on labor flows. Walmsley and Winters (2005) (Table 4, col. V) usee a
general equilibrium model to estimate that allowing an additional movement people
equal to 3% of the existing OECD labor forcewould raise the welfare of those moving
by $170 billion.8 Again, the logic is familiar, since welfare gains grow with the square of
the deviation and goods existing price distortions are measured in percents and labor
price distortions are measured in factor multiples, the gains from labor mobility just
swamp everything else on the agenda. Figure 5 shows these gains from complete labor,
goods, or capital liberality on the same scale, as, as you might suspect it is impossible to
see the gains to further flattening from the spectacularly high cliff of 65 trillion.
Figure 5: The view of the flat land of goods and capital from the cliff against labor
70000

65000

billions $

60000
50000
40000
30000
20000
10000

65

109

Complete
equalizatoin MPK

Complete
liberalization of goods
markets, to
developing countries

0
Complete
liberalization labor
mobility

Source: Hamilton and Whalley (1984) and Klein and Ventura (2004) for labor
mobility, Caselli and Freyer (2006) for capital, World Bank (2005) for goods.

Their simulations were based on rough assumptions about how much of the existing wage differences were due to
productivity differences that would move with the worker (which they assume is only half) while our estimates per
worker are based on data. They find a welfare gain to movers of roughly $20,000 per mover, which is close to our
estimate for India of P$19,900 (though their estimates are not exclusively of low skill workers).

The world ran an interesting and unique historical experiment since World War II
of dividing the world up into smaller and smaller geographic bits, endowing those bits
with sovereignty over economic policies, and then promoting modest amounts of crossborder liberality in some transactions (certainly goods, certainly foreign exchange, less so
for capital) but almost uniformly countries block the movement of labor9. Different
simple theories of economic growth and/or international trade made different predictions
for how that experiment would turn out, some predicting convergence in per capita
incomes across countries, some predicting equalization of factor prices. We have now
run with the experiment with the POSEBLL for around 60 years and we know the
outcome: incomes have not converged (certainly not in absolute terms, certainly not in
country weighted relative terms) and factor prices have not converged. We dont know
what real globalization might have produced but POSEBLL has created a world in
which there are massive differences in the earnings of equal intrinsic productivity
workers across countries, differences that are sustained by enforcing restrictions on the
movement of poor people.
III)

The Nation-State-ization of Equity

One issue that is rarely raised in discussions of globalization and equity is what
the concepts of equity or fairness or justice mean in a global or even cross-national
context. While I am not well suited or trained to raise these issues, if not me, then who?
One of the hubs, if not temples, of globalization is the Singapore Airport. Nearly
every experienced international traveler has passed through it. It is a marvel of
9

What is unique is that most, if not all, countries blocked the influx of labor. While the movement of labor
across state borders has always been complex, states in previous times were at least as concerned with
losing labor (hence all kinds of arrangements from serfdom to slavery to peasantry) that bound people to
the land to prevent losses of labor (especially in rural areas), but much less concern about cross-state
mobility.

modernity, efficiency, a testament to non-Western wealth and prosperity, and a hive of

globalization as business-people from every corner of the globe pass by. A few years
back I was headed towards my next flight when I encountered a string of Bangladeshi
men; all handcuffed and then chained together being escorted through the airport,
presumably to be flown back to Bangladesh. Their apparent crime was an attempt to
sell their labor services to willing buyers in Singapore. None of us streaming past paid
the slightest attention to this perfectly natural, perfectly ordinary, course of events.
One fascinating aspect of the huge divergence in earnings across national borders
is that it takes coercion to enforce those restrictions. This coercion, at least in the OECD,
is carried out by agents under the near perfect control of democratic nation-states. There
is overwhelming support for these restrictions, in nearly all public opinion surveys in
OECD countries recently (even before the troubles in 2008) a majority of people believe
that, even with the existing restrictions there is still too much migration. These
restrictions raise almost no objections on the grounds of justice or equity. This is
apparently because of what I call the nation-state-ilization of justice claims.
I would argue a desire for equality often stems from an even deeper principle of
equity, that likes should be treated like likes. Even a very small child will object
thats not fair if any favor is distributed unevenly if that uneven distribution appears to
be arbitrary. However, what constitute like for purposes of justice claims, is, as they
say, socially constructed, as what differences count as making people unlike is entirely
a social convention.
For instance, there is little that is more obvious about the world than that the
biological sexes differ. Nearly every individual is immediately and easily recognized by

all others as belonging to one or the other of the biological sexes. Yet, in most modern
societies these obvious differences are now irrelevant for justice claims, the differences
between the sexes have been redefined as irrelevant to justice claims, as social
constructed notions were deconstructed and reconstructed. Because you are a girl is no
longer considered a socially appropriate rationale for differential treatment.
On the other hand, people who are exactly identical in every conceivable and
observable respect can be treated in ways that cause their well-being to differ by orders of
magnitudefor instance denied access to a more productive job--with no apparent
violation of justice if those otherwise identical individuals happen to be citizens of
different countries. Two brothers both born to, say, Peruvian parents, if one is physically
born in the USA and one not, have completely different lifetime claims on rights as this
seemingly arbitrary condition of place of birth acquires makes them completely unlike
for nationalized theories of justice.
Just as one illustration, we can use our multi-country wage data to ask the
question of how apparent discrimination against women in wages in the labor market
just simple wage differences between otherwise observable men and womencompare
with the wage gap between observably identical men (including remember same country
of birth). Not surprisingly, the estimates are consistent with the existence of substantial
labor market discrimination against women in nearly every country of the world. In the
USA the estimate is a wage ratio (men to women of 1.3). Estimating the male
premium for each of the countries gives a median of 1.4 (shown by the median country,
Madagascar, in Figure 4). The worst observed male premium is, again not very
surprisingly, in Pakistan, with a male premium of 3.1makes making three times

observably equivalent females. Of course these are very simple numbers just meant to be
illustrative and are not corrected for labor market selection, nor do these reflect the many
potential dimensions of sex discrimination beyond the labor market (e.g. violence against
women, including domestic abuse, which is widespread in many countries, forced
marriages, bias in health care, property ownership, etc.). But the simple point is that our
very conservative estimate of the wage gap of equal productivity workers willing to move
(taking observed wage ratios for observably identical workers and dividing by 1.5)
exceeds the median male premium (1.4) in nearly every instance (only two countries are
less than this amount), and 17 countries the place premium to access to the US labor
market for low-skilled workers is greater than the worst discrimination observed in any
country10.

10

These numbers are derived from a modestly difference technique and functional form than those in table
1 and hence are not completely comparable.

Figure 6: Comparing wage gaps across borders (estimates from table 1, scaled back
by 1.5) to wage gaps within spatially-integrated labor markets
9

8.4

Domestic wage ratios between observably identical members

of different groups

8
7

Wage ratios between observably identical workers in spatially

integrated labor markets

Selected cross-country estimated wage ratios for observably

and unobservably identical workers, R e = R o /1.5

4.9

5
4
3.1

3.2

2.7

3
2
1.3
1.1 1.25

1.1

1.3

1.4

1.4

1.4

1.4

1.4

1.5

1.5

1.6

1.9

1.9

M
95
ex
ic
.U
o
v.
O
1
S
99
th
M
st
er
9
al
at
ay
v.
es
M
,M
In
1
en
di
99
al
ge
v.
ay
9
no
W
si
a
u
o
s,
m
U
1
98
en
rb
Bo
an
8
,M
liv
v.
ia
ad
ru
19
a
ga
ra
89
l,
sc
43
ar
co
20
U
S
un
01
v.
tri
Pu
es
er
19
to
99
R
ic
U
o
W
S
19
As
hi
v.
99
te
ia
Ita
n
v.
l
y
v.
Bl
,1
Af
ac
92
ric
k,
1
an
U
S
,T
19
an
39
za
U
S
ni
v.
a
19
Ja
M
71
m
en
ai
U
ca
S
v.
v.
19
W
Bo
om
99
liv
en
ia
,P
19
ak
99
is
N
t
ew
a
U
n
S
Ze
20
v.
01
al
In
an
di
d
a
v.
19
To
99
ng
U
S
a,
v.
20
N
05
ig
er
ia
19
99

io

bt

v.
S
U

se

ne

hi

ax

.r
at

en

v.

om

en

,U
S

19

19

76
ne
ic
ro
M

v.
m

i/G
ua

ai
aw
H

si
a

19

95
19

ia
nd

,U
S

s,
I

ck
as

te

Bl
a

.c

v.

ed

hi
te

Sc
h

v.
rs
th
e
O

97

Source: CMP (2008a), Figure 4.

One can do similar comparisons with other forms of labor market discrimination.
One study estimated the degree of wage discrimination against African-Americans in
1939a time in America where discrimination was egregious, blatant, and pervasive.
The estimate was that the wage ratio was 1.6, whites made 60 percent more than
equivalent African Americans. This was a clear offense to any sense of justice, and many
people fought long and courageously to address that, and current estimates suggest much
lower degrees of outright wage discrimination. But, even the outrageous discrimination
of 1.6 is exceeded again by all but two countries (Mexicos estimate is 1.61). Estimates
of labor market discrimination against disfavored ethnicities around the world, e.g. versus

indigenous groups in Bolivia, often show persistent and large wage gapsbut never
reaching the levels of the discrimination of the US border.
The question is, how does the massive differential treatment of people that are
alike in every respect except for their affiliation with a particular nation-state, an
essentially arbitrary condition of birth, square with any theory of justice? This is not to
ask why there are not in fact open (or more liberal) policies with regard to labor mobility
(or even deeper, the acquisition of citizenship), which is a historical and political
question, but rather why it has come to be accepted that this differential treatment of
like individuals is not a violation of fundamental principles of equity. Alternatively,
what is the construction of a notion of difference in a theory of justice such that nearly
all other conditions of birthe.g. sex, race and non-biologically grounded socially
transmitted ascriptive identities, such as ethnicity and religionare absolutely
unacceptable as criteria of difference, particularly for action by the state, but your
location of birth (or parentage) is an acceptable basis for legally grounded and coercively
enforced discrimination by states?
I will be the first to admit I am a puzzled amateur rather than a professional
philosopher. To me the three primary theories of justice, when applied to this issue either
imply open borders, or, in an effort to avoid this obvious conclusion, devolve into a mass
of confusion or irrelevance.
One version of a popular class of theories of justice (contractarian) is that
articulated by Rawls in his classic Theory of Justice. His basic argument has two parts,
first, that a set of social arrangements should be considered just, if individuals behind a
veil of ignorance and hence with no knowledge of the position they would

subsequently occupy in that social arrangement, would have agreed to those social
arrangements, and second, he makes arguments about what social conditions would in
fact be agreed to in those circumstances. The seemingly obvious implication of that setup is that open access to participation to any given set of social arrangements (conditional
on fulfilling whatever obligations those arrangements entail) is a fundamental condition
of justice, and hence open borders one corollary of that. Imagine, taking Rawls perhaps a
bit more literally than he would prefer, we actually existed in some sense as entities
capable of reasoning before birth, say, raw intelligences. In the pre-birth conference of
intelligences would I ever agree that you would be born in Denmark or France or the
USA and I would be born in Mali or Nepal or Bolivia and that it would be just for you to
arrange for coercion to prevent me from working in the territory controlled by your social
arrangement, even for a willing employer? Hard to see why I would. Philosophers such
as Joseph Carens argue that Rawls approach does imply open borders.
Rawls own means of avoiding this consequence of his approach (which may be
seen as a defect as some might regard any theory of justice that implied open borders as
flawed as it would imply to modern nation-state was just) was simply to stipulate the
nation-state as a primary and primordial entity and claim his theory only applied to social
arrangements up to the nation-state and no larger. This is a radical reduction of the scope
of his theory, as it means to have any general human or universal theory of justice as we
must have one theory, Rawls say, for relationships within a nation-state (a potentially
arbitrarily formed category) and another theory, which cannot be Rawlsian but is
otherwise unspecified, to answer questions about the just relationships between
individuals across nation-states and perhaps a third for behavior of nation-states as actors.

For instance, are there any conditions for determining new members of a nation-state that
are unjust? For instance, could a nation-state deny a person admission as a citizen on
precisely the same criteria (e.g. gender, religion) for which discrimination against
national is unjust? This seems like a mass of confusion to avoid a simple, plausible,
universal, veil of ignorance contractarian theory and its obvious consequence.
The philosopher Robert Nozick proposed a comprehensive alternative view of
justice as process fairnessthat people own the fruits of their labor and are entitled to
make voluntary transactions and that any outcome that results such a process is fair, not
matter how unequal. This theory as well leads quickly to the conclusion that it would be
unjust to bar voluntary transactions across individuals on the basis on some justicearbitrary condition like their place of birth. The only alternative is to again introduce the
nation-state as an unanalyzed primordial with an ability to regulate or ban transactions
between citizens of nation-states it would be unjust to bar for citizens of the same nationstate.
A third alternative class of theories of justice are more communitarian and build
notions of justice up from sustained free dialogue among the participants of a community.
These theories, while having attractions, seem especially problematic as the conditions
under which individuals can join communities and the ability of communities to regulate
membership seem particularly intractable. If justice is only community based then is
there any injustice to a community denying new members based on, say, their race? If
so, then there must be some other theory of justice that transcends the community, a
broader community notion of justice. The problem with this sequence is that there is no
reason why this process of envisioning broader communities would just happen to stop at

the (imagined) community called the nation-state which could then deny any and all
others membership without any justice claim. There are some, perhaps many, things that
all human beings have in common, which would seem to imply some set of universal set
of justice obligations and a common notion of equity such that, while communities
may have distinct notions of justice, there are at least some cross-community constraints
on the range of acceptable cross-community actions consistent with fairness, equity, or
justice.
Of course the alternative to these views is the widespread view, of which Foucault
is the most popular well-spring, that, very crudely put, discourse is structured by power
and that what discussions that pretend to rational discourse about all topics, including
justice, are the cloaked attempt of power to control discourse to construct a social reality
conducive to its aims. The alternative to this discourse is deconstruction, the
unmasking the relations of power behind discourse so that alternate realities may emerge.
However, this post-modern deconstruction is unlike most previous deconstructions in
that it has given up any illusions of displacing the false with the truththere is not
solid, much less, transcendent (in the sense of God or Reason or History) basis for
truth or justice it is just discourse and convention all the way down. Hence postmodern analysis is not an alternative theory of justice, but rather a theory of justiceclaims.
Taking Foucault seriously then one should look for power not in controversy but
in silences. After all, controversy reveals fissures in power, and signifies either a topic
beyond the real concern of power, an erosion of power, and/or conflicting powers in
which neither is able to subjugate the other. In issues around which there is

overwhelming power there is no legitimate controversy, only silence, because discourse

beyond the specified boundaries is not controversy but craziness. This seems like a
good description of the current status of justice discussions about cross-border movement
of people. The claim that closed borders are unjust isnt controversial, it is just plain
crazy.
This lack of a widely acceptable notion of cross nation-state border equity leads
the discussion of the movement of people into twists and knots, let me illustrate three.
First, I wrote part of this while in Kuwait. Kuwait has a massive population of
non-Kuwaitis, most of whom are explicitly temporary workers and most of whom are not
and are not on a path to achieve any long-term claims on citizenship in Kuwait. Kuwaitis
and the government of Kuwait seem perfectly comfortable with co-residence on unequal
citizenship status. One can articulate a view, based on Nozick view, that this is a just
arrangement: We make an offer of the conditions under which people may be granted
access to the Kuwait labor market (including their limits on rights, not being able to bring
families, less than due process for expulsion, etc.) and if people accept that offer then by
revealed preference they are better off, so by process fairness the resulting outcome is
just (or at least no less just than the situation before we made the offer and that initial
injustice is not our problem). Many people are uncomfortable with this view, as they
argue that it will lead to a race to the bottom or a coarsening of a sense of justice
such that within national inequality will grow or Kuwait will be unable to sustain social
programs. For instance, Milton Friedman himself proclaimed that the welfare state and
open migration are incompatible. As a general proposition this is obviously refuted by
Kuwait (and other Gulf States and Singapore) all of which, for their citizens, have an

amazing cradle to grave set of social welfare programs, often even beyond most
European states. What Milton Friedman meant was that social welfare states and open
migration are incompatible if migrants acquire immediate and full claims on these
benefits. The conundrum is that the Kuwaitis lack of a sense that co-residence creates
justice claims leads them to accept a far higher number of workers than they would if
each acquired citizenship (which of course implies a claim on the revenues from oil).
The lack of a coherent theory of cross-border equity means there is not coherent view
about whether allowing temporary workers is more or less just than banning them
altogether.
A second conundrum is that the conventional wisdom is that a theory of universal
human rights does involve allowing certain kinds of petitionslike refugee or
asylum but a nation-state having no justice obligation whatsoever with regard to
economically motivated migration. Suppose, an asylum petitioner could prove that
their odds of dying if they were forced to return to their country were one in five within
five years. This is a fantastically high fatality ratefor comparison the fatality rate of the
US Army in World War II was 2.8 percent, the British Army 5.2 percent. Almost if a
one in five fatality risk was demonstrated in asylum hearings their petition would be
granted. In the poorest countries of the world the Under Five Mortality rate is around
200 per thousandone in five. So returning a Malian woman or Somali or Liberian
woman who desires to remain on mere economic grounds creates this incredibly high
mortality risk for all of their future children. Why isnt every Southern Sudanese or
Somali or Malawian or Nepalese a candidate for asylum, not from political prosecution
but from the very real dangers of poverty?

A third ethical problem is that ambivalence of Americans about the justice of our
current situation. Many, if not most, Americans do not want open borders but feel little
or no ethical compunctions about violating the current immigration laws by employing
undocumented workers. The most recent, spectacularly illustrative example is the recent
arrest (December 2008) in the Boston area of the top local official of Customs and Border
Patrol for not just hiring an illegal worker, but actively aiding this person in evading the
law11. This deep ambivalence I would argue stems from the conflict between the vague
sense it is just to enforce borders competing with the similar vague sense it is not
unethical to give another person work.
All in all, I find myself confused and out of touch in discussions about
globalization and equity as I do believe in God but dont believe in Sudan, whereas
everyone around me seems to have the opposite view. As Benedict Anderson so cogently
pointed out in Imagined Communities nationalism is the last acceptable credo and the
imagined communities of nationalism have swept the floor with not just kith and clan
but God and Class. The puzzle is not that people believe in states, these are an obvious
juridical category, the puzzle is that people believe in nations and hence nationality as a
social category on which discrimination is possible without any justice claims. Yet
social realities of the Congo or Indonesia or even India are communities or
nations that, like Yugoslavia or Sudan, exist only in the most fevered of imaginations.
For me, Ill take Pascals wager and stick a while longer to the old fashioned notion that
all men and women are brothers and that a theory of justice should be universal, not
contingent on place of birth or physical proximity or nationality.

11

Ironically, as I arrived in February of 2009 her picture still greets you as I cleared Customs.

Conclusion
My addition, or perhaps subtraction, to a discussion about globalization and
equity is to triply deny the premise.
First, for most poor countries globalization is a much less primary phenomenon
than sovereignty, and even the most aggressively liberalizing countries have yet to
overcome the dis-integration consequences of sovereignty.
Second, the range of equity issues usually discussed in this context are all
dwarfed, by order of magnitude, by the wage gaps across equally productive workers
created by the enforcement of US (and other rich country) borders. Given that the typical
low skill worker could triple his/(her) wages by moving to the USA (or other rich
countries) discussing the inequity of cotton subsidies or fair trade or inadequate foreign
aid, or the efficacy of anti-poverty interventions leaves one wondering, why this and not
that?
Third, I cannot see there is any coherent way for discussing equity (or fairness
or justice) in the context of globalization that doesnt seem to boil down to an
unsupported claim about the primacy of nation-states as a legitimate ethical category.
To me this seems more like a tradition of all dead generations [that] weighs like a
nightmare on the brains of the living from which I suspect we may one day awake.